💾 Archived View for thrig.me › blog › 2023 › 11 › 24 › lag.gmi captured on 2024-07-09 at 01:14:13. Gemini links have been rewritten to link to archived content
⬅️ Previous capture (2023-12-28)
-=-=-=-=-=-=-
So there was the usual discussion of colors in editors—no, I'm not enabling them. A novelty this time was that someone suggested vim, which I had already dropped for vi some years ago. Maybe they thought an even more colorful IDE was being used? Apparently Visual Studio is real bad. My reasons for dropping vim include the slow startup time (recall the 2009 MacBook with the spinny disk), sluggish cursor motions, not to mention having to disable various new things they kept adding, in addition to dealing with wacky vendorisms. (vi is also bloated, but ed lacks filters and some other nifty things ex added, if you want to argue to the standard editor.) Shouldn't software be getting faster and more efficient over time?
I've had this nagging feeling that the computers I use today feel slower than the computers I used as a kid. As a rule, I don’t trust this kind of feeling because human perception has been shown to be unreliable in empirical studies, so I carried around a high-speed camera and measured the response latency of devices I’ve run into in the past few months.
If we look at overall results, the fastest machines are ancient. Newer machines are all over the place. Fancy gaming rigs with unusually high refresh-rate displays are almost competitive with machines from the late 70s and early 80s
The "Doherty Threshold" clocks in at about 400 milliseconds; above this the user interface experience is apparently too slow. A counterpoint is that 400ms of lag in music would be terrible. With practice such lag might become tolerable, say on a pipe organ, though you'd probably end up playing from a mental and physical state and ignore the outdated and conflicting auditory information, similar to typing on a laggy computer with your eyes closed. (About that typo rate.) Anyways, there are likely different expectations for computer interfaces and music; live music production is in general much more sensitive to lag. Even 40ms of lag could be real bad.
Contrapunctus VIII by J.S. Bach
Particularly dreadful was an embedded Windows point-and-shoot camera; it took… a while… to start, and you would often take a photo of your shoes because you would press the button, wonder what the camera was doing, tilt it down to look at the display, and then it took the photo.
Conversations can probably tolerate more lag than music (a human may need some time to get a thought together, but not too long) so maybe that's where the 400ms comes from?
Worse is when computers are inconsistently laggy; a pipe organ usually does not randomly decide that throughput is more important than latency. General purpose operating systems have to balance those two, somehow. Modern systems also tend towards the accumulation of too many layers along the key-to-display pipeline, which Dan Luu touches on. And it gets worse when a network is involved…